Current network infrastructure design is based on the assumption that devices may be fixed or mobile, but the equipment that provides wireless services to these devices (e.g., routers, cable modem termination systems, digital subscriber line access multiplexers, or wireless base stations) are generally fixed. Thus, there is an inherent maximum amount of network capacity (measured in upstream and/or downstream bits per second) available to the devices. Further, some devices may be subscribed to a wireless service provider according to a plan that supports a maximum upstream and/or downstream rate that is capped at less than this inherent maximum amount.
Nonetheless, data transmissions are often bursty—a device may use up to 100% of its available capacity for a short period of time, but otherwise remain relatively idle. As a consequence, the device is not able to obtain full utilization of its network capacity. Indeed, many network service providers overprovision their subscribers, resulting in much of their network capacity being used inefficiently. Nonetheless, in some cases, the device would benefit from more capacity than is available.